2025-07-08 ACCESS EC Meeting Agenda & Summary

(v) indicates a voting member

2. Quarterly Meeting

   1. Drop Dead Info: September 2025 ACCESS Quarterly Meeting - PlanningPreview

      1. John: no further info on the supplement from U of I - August 6 is the earliest date

         1. Stephen - PSC has accepted some with the new amendments

         2. Amit - SDSC is accepting some

         3. John will reach out and talk to Amit and Stephen for details

   2. Will begin September meeting planning soon

      1. Who will be taking Lavanya’s place on the agenda planning meetings?

      2. Cindy W is taking the reins - there is a meeting next week

3. ACCESS Booth at PEARC - need more volunteers to staff the booth (Chuck)

   1. https://docs.google.com/spreadsheets/d/1pm60p9dwYbvNOOA3cxcFV0QAwdQP6JlWpti5w1hx_Eg/edit?gid=0#gid=0

   2. remind teams please

4. What/Who is ACCESS - Shelley (30 mins)

   1. may not finish conversation today

   2. prompted by NSF recommendations:

      1. 1st item on the list: Discussion is needed among the five awards and agreement needed on
         what type of resources should be integrated* into ACCESS. Note that all ACSS-funded resources need to be integrated into ACCESS by definition. Other types of aMiliation of resources and organizations with ACCESS need to be discussed by the five awards and agreed upon by the EC, with a clear benefit to the user community for the cost/effort involved with the aMiliation. “Integrated resource” vs “aMiliated resource” needs to be clearly defined, differentiated, and presented by
         ACCESS (on the ACCESS website, for example), to avoid confusing people looking for resources.

         • The term “Integration” generally means the resource is an allocatable
           resource with an allocatable (by ACCESS) capacity that makes the
           cost/benefit of integration and allocation beneficial.

   3. are we for everyone or some and not others - and who are those people

      1. Tim - we are described as what we are and who we serve in the solicitation - we are meant to support the NSF cyberinfrastructure ecosystem

         1. how we prioritize our limited resources is a different discussion

         2. if the scope is going to change from the solicitation - it needs to come from NSF and they should define the changes

         3. This is from the standpoint of resources

      2. Stephen - “are you NSF funded is the first step in engaging priority”

         1. no we do not serve everybody - this is what the administration is telling us

         2. there are certain domains and types of research we are no longer allowed to support

         3. NSF charge - we need to describe the people we serve better

         4. We should capture the most people and better target them - Go thru statekolders and prioritize them so we can make engagement and communication targeted to them

         5. but we are certainly not ACCESS for all anymore

         6. This is standpoint of users

      3. John - we have two points - what are we and what is the community we support

         1. What are we and who we serve is covered by the solicitation

         2. This is modified by government in power

         3. but it has not been clarified by NSF - and we need this

         4. This makes things tricky - we are not prohibited from serving certain communities - but we are prohibited from singling them out for priority access

         5. We need to equitably serve - but this is not allowed under current administration - we now have to equally serve

      4. Shelley -

         1. who comprises the team serving the people we serve

            1. this we know

         2. What do we do beyond that

            1. put users on the shelf for a moment

            2. Do we only want to serve the RPs funded thru ACSS - and people who have NSF funding?

            3. Do we go beyond this or do we stick with this

               1. Tim - we should go beyond - it is part of our solicitation and is part of other solicitations as well (CSSI contract includes being served by ACCESS)

         3. If we consider people other that those funded ACSS - what do we serve them and at what priority?

            1. They are not just required to participate in ACCESS - they are required to participate in a particular way - others are not required to participate in a particular way - this is one difference

            2. ACCESS is a child project from XSEDE and other programs going back in time

      5. John - it used to be that there was an expectation that OAC awards had to also participate with ACCESS

         1. not sure if this is still being discussed

      6. Puri - we are now in a different world - and what ever conversations have happened in the past are in the past

         1. Cloudbank will be part of ACCESS in a couple months - should we be focusing on making this part of ACCESS

         2. There is the comparison with a CSTAR resource which may only have 4 users

         3. In order for ACCESS to be successful and move forward - we need to be able to show growth

            1. priority and effort should go to Cloudbank - not because they are less important but because the future will be reaching out to the larger audience

      7. Tim - we will continue to be successful and continue moving forward - our success will not be linked to something we are

         1. we are maintaining existing operations - there is only a small slice of time that can be focused on new things - 60 to 70% of our effort is moving forward with inflight ACSS resources

      8. Tom - if we dilute our services with more - we will not be able to serve what we have well

         1. we can serve without integrating

      9. Solicitation

         1. “ACCESS is an NSF-funded program that provides the U.S. research and education community with coordinated access to a nationwide ecosystem of advanced computing, data, storage, and other cyberinfrastructure resources–at no cost, and without the need for grant funding. It serves as a central hub connecting systems, expert support, and training to accelerate scientific discovery, innovation, and education across all disciplines.”

            Hearing this discussion makes me think that the ACCESS “central hub” also connects other NSF programs - which is missing from the current language.

         2. “ NSF encourages exploration of various delivery mechanisms, including but not limited to, those leveraging the Advanced Cyberinfrastructure Coordination Ecosystem of Services & Support (ACCESS program)”

            1. https://www.nsf.gov/funding/opportunities/cssi-cyberinfrastructure-sustained-scientific-innovation/nsf22-632/solicitation

      10. John - Puri’s guidance is different that what we got a year ago - we need a defininative guidance in order to do prioritization - NSFs priorities have changed and this has not been communicated well by NSF

          1. Tim: That's exactly the point I made in my first comments at the start of the discussion. If we are to change scope, we need that from NSF.  We also need NSF to communicate to the broader community that they are changing our scope.

      11. Puri - we will not get any more guidance than what is on the website

      12. So should we just say - we will follow the guidance in our solicitation

      13. Shelly - shoring up the groups we are committed to service before we start diluting into doing everything for everybody is our best focus - serve the people we need to serve properly first

      14. Tim - ACSS is mandated to integrate with ACCESS up thru allocation - but some things are optional for them - but we don't' have the stick to get them to participate - only NSF has the stick to tell them to do it

      15. John - based on what guidance we think we have - we should define what it is we are and who we serve - we will provide that to NSF - and then we will probably get feedback for changes - which could all be avoided if we had guidance first - but we may not get it

          1. Puri - the formal notes from the review are the specific guidance

      16. Tom - we know our systems better than NSF - let us say what we can / cannot do - how well are we providing service to end users

      17. John - we have cooperative agreements with create flexible opportunities to define what we do - which would still have to be provided to NSF to approve - do we really want to redefine what we are 3.5 years into a project - we can ID a core set

      18. Tom - this is the edge cases at this point

      19. From Sharon: "Discussion is needed among the five awards and agreement needed on what type of resources should be integrated* into ACCESS. Note that all ACSS-funded resources need to be integrated into ACCESS by definition. Other types of affiliation of resources and organizations with ACCESS need to be discussed by the five awards and agreed upon by the EC, with a clear benefit to the user community for the cost/effort involved with the affiliation. “Integrated resource” vs “affiliated resource” needs to be clearly defined, differentiated, and presented by ACCESS (on the ACCESS website, for example), to avoid confusing people looking for resources."

      20. Shelley - we need to define our priorities

      21. Stephen - ACSS category 1 and 2 - good start – are we doing a stellar job there? are we supporting as well as we can?

          1. we don’t need to go to MRI or CCSTAR

          2. stick with our bread and butter and worry about the rest after we do better for our main users

5. “User Like Me” Resource Recommender Model - Tom (10 mins)

   2. what his team is working on

   3. if it is liked by the EC they can form a user group

   4. Puri - would like to be invited to the initial user group meeting

   5. Tom and Stephen are interested in leading a new WG

   6. Chuck and Shannon will help get the group started

   7. At PEARC get together with developers on teams to discuss - and then form a working group

Program Milestones

Open

EC Meeting Task Tracking + Quarterly and Annual Meetings To Do Follow Ups

Open

Misc Topics:

Next EC Meeting: 15th July 2025, will there be a meeting July 22? (no)

Reference:

Risk Register: https://access-ci.atlassian.net/jira/core/projects/RR/board

Project Change Request: https://access-ci.atlassian.net/jira/core/projects/PCR/board